Good to Great

Collins and a research team spent five years looking for companies that had been merely good for a long stretch and then sustained fifteen years of exceptional returns, and comparing each against a rival that never made the jump. Eleven companies qualified. The book is what the pairs had that the others didn't.

The findings are the vocabulary that entered every boardroom afterwards: Level 5 leadership, where the chief executive is personally modest and professionally relentless; first who, then what — get the right people aboard before deciding where to drive; the Hedgehog Concept, the intersection of what you can be best at, what drives your economics, and what you care about; and the flywheel, which is the argument that transformations look sudden from outside and never are.

It is worth reading alongside the objection, which is that picking winners and then finding their common traits is selection on the dependent variable, and that several of the eleven later did badly. The framework is more useful than the sample, and the flywheel chapter in particular has outlived the study it came from.